AI-native product studio · India
We build AI automation and software that
Agents, internal tools and product builds — scoped against a reference you can click before you sign, then shipped weekly into a repository you own.
- Around ten people, a few builds at a time
- Nine applications you can open right now
- One published method, versioned and dated
No headline figure. Counts you can check.
A studio's proof is usually one big unverifiable number. Ours is deliberately segmented — what has been delivered, and what you can actually open, kept apart and never added together. Every figure below is a link to the thing it counts.
What that looks like publicly: 9 of them run, 27 are navigable screen sets with no live backend, and two more counts sit alongside them.
- 9runnable applicationsFull applications you can open and click through end to end.Check it
- 27reference buildsNavigable rebuilds of delivered systems. Real screens and flows, no live backend.Check it
- 8native design buildsiOS and Android screen sets. Designed and specified, not compiled apps.Check it
- 6industry studiosPer-industry collections of individually designed sites.Check it
The usual engagement, and the one we actually run.
Four of the eight rows we publish. Every row on our side names the cost that comes with it, because a comparison where one column has no downside is a brochure.
Who you actually talk to
An account manager relays your questions.
The engineers writing the code, in a shared channel.
How scope gets agreed
A long document, read two different ways by week four.
You click a reference build first, then we write the scope around it.
What you get each week
A status deck and a percentage that only goes up.
A deployed build on a staging URL, and what is blocked.
Who owns the code
Ownership transfers on final payment.
Your repository from the first commit.
Most studios trademark their method. We published ours.
Most studios name a method, put a trademark on it, and keep the detail behind a sales call. We publish ours instead. Read it before you talk to us, argue with the parts you disagree with, and hold us to the rest.
Reference-First Engineering · v1.0 · revised 26 July 2026
A reference build comes before the estimate
We do not quote a system we have not built a version of.
The people who scope it write it
No account layer, no handoff from the people who sold the work.
Every week ends with something deployed
Progress is a URL you can open, not a percentage on a slide.
Decisions are written down next to the code
Every non-obvious choice leaves a paper trail in your repository.
Model-backed features start with the test set
We agree how a feature will be judged before we build it.
Three shapes of engagement, each with a stated minimum.
No tiers and no packages. Three shapes, and for each one the smallest thing we will actually start — so you know what you are agreeing to before anyone writes a scope.
A scoped build
One system, a recognisable finish line, a handover at the end.
Smallest startThe reference build stage. That is the smallest thing we will start, because starting a full build without one is how projects go wrong.
An ongoing partnership
A standing allocation of senior time against a roadmap that keeps moving.
Smallest startOne full cycle. We do not take partnership work in fragments, because the ramp-in cost lands on you.
Advisory
Senior review for teams who have their own engineers.
Smallest startA short block of sessions rather than a single call. One conversation rarely produces anything your team can act on.
3 service lines. Not seven.
Seven near-identical pages read as SEO surface. Three with real depth is what we can stand behind — each one says what it covers, how it gets scoped, and which reference builds demonstrate it.
AI agents & automation
Agents and automations scoped to one workflow at a time, with evaluation and human review designed in from the start.
What this coversMVP & product builds
Zero-to-one product builds where the deliverable is a running system plus everything needed to own it.
What this coversInternal tools & ops
Dashboards, back-office systems and operational tooling for teams whose real process lives in spreadsheets and inboxes.
What this covers
The same three lines, by sector:Logistics & MobilityReal Estate & PropTechHealthcare & MedTechManufacturing & Industrial VisionFintech & Finance OperationsSaaS & Customer SupportLegal TeamsAccounting, Tax & BookkeepingAutomotive Retail & AftermarketConstruction & ContractingEducation & EdTechField Service & TradesInsurance & ClaimsMarketing, Advertising & AdTechMarketplaces & Two-Sided PlatformsMedia, Publishing & StreamingRestaurants & Food ServiceRetail & E-commerceSports, Fitness & AthleticsTravel & Tourism
Three you can open right now.
Client repositories are private, so each of these is our own reference build of the delivered system, on sample data. These three are full applications rather than screen sets — open one and click through it end to end.
Runnable applicationBrief Forge
Boutique law firm (NDA)
Contract review AI for solo lawyers and small firms — extract, score, and redline contracts in minutes.
Runnable applicationPatient Front Desk
Multi-clinic healthcare group (NDA)
AI receptionist for clinics — auto intake, smart scheduling, and insurance verification before the patient walks in.
Runnable applicationSupport Pulse
B2B SaaS support team (NDA)
AI ticket triage and drafted replies for SaaS support teams — cut first response time and stop SLA leaks.
6 industry studios. 128 sites, designed one at a time.
These are marketing sites, not software, and they exist to show design range across categories at a volume that is hard to fake. 100 of them are for brands we invented. The other 28 are unsolicited concept revamps of real businesses that never commissioned, briefed, approved or paid for the work — which is stated on the studios page too.
Open the studio program- The Pass StudioCafés & F&B · 20 sites
- The Rep StudioGyms & Fitness · 20 sites
- The Close StudioReal Estate · 20 sites
- The Docket StudioLegal & Law Firms · 20 sites
- The Rounds StudioClinics & Healthcare · 20 sites
- Atelier Travel StudioTravel & Tourism · 28 sites
Stop reading proposals.
Start clicking the build.
Send us what exists — a running product, a repository, a half-finished build, or just the constraints. You get a written read of it back: what is solid, what breaks under change, and what we would cut. You keep that whether or not we work together.
- No long-term contract to start. The reference build is a bounded stage on its own.
- Work lands in a repository you own from the first commit, not in a handover zip.
- You talk to the engineers writing the code, in a shared channel, not to an account manager.
The honest constraint: replies and support run business hours, Monday to Friday, IST, and we take on few builds at a time. If the calendar does not fit, we will say so at the build review rather than after you sign.